用图结构组织技能，让大模型智能选、组合、执行任务。
GraSP: Graph-Structured Skill Compositions for LLM Agents
- 将技能构建成有因果关系的有向无环图，实现精准调度。
- 在多个任务上提升最高19分奖励，减少41%环境步数。
- 适合需要复杂决策的大模型代理，尤其抗过量技能干扰。
大模型代理的技能生态快速发展，但新基准显示：技能越多表现并不越好——2-3个精选技能反而优于全面文档，过多技能甚至损害性能。瓶颈已从技能数量转向技能编排：代理不需要更多技能，而需要一种结构化机制来显式选择、组合并执行技能。我们提出GraSP，首个可执行的技能图架构，引入技能检索与执行之间的编译层。GraSP将扁平技能集转化为带类型标注的有向无环图（DAG），通过前提-效果边表达因果依赖，执行时进行节点级验证，并利用五种类型的操作符实现局部修复，将重规划复杂度从O(N)降至O(d^h)。在ALFWorld、ScienceWorld、WebShop和InterCode四个数据集上，使用八种LLM骨干网络，GraSP在所有配置中均超越ReAct、Reflexion、ExpeL及扁平技能基线，最高提升19分奖励，环境步数减少41%。其优势随任务复杂度增加，且对技能过量检索和质量下降具有鲁棒性，证实结构化编排才是可靠代理执行的关键。
原文摘要 · Abstract (English)
Skill ecosystems for LLM agents have matured rapidly, yet recent benchmarks show that providing agents with more skills does not monotonically improve performance -- focused sets of 2-3 skills outperform comprehensive documentation, and excessive skills actually hurt. The bottleneck has shifted from skill availability to skill orchestration: agents need not more skills, but a structural mechanism to select, compose, and execute them with explicit causal dependencies. We propose GraSP, the first executable skill graph architecture that introduces a compilation layer between skill retrieval and execution. GraSP transforms flat skill sets into typed directed acyclic graphs (DAGs) with precondition-effect edges, executes them with node-level verification, and performs locality-bounded repair through five typed operators -- reducing replanning from O(N) to O(d^h). Across ALFWorld, ScienceWorld, WebShop, and InterCode with eight LLM backbones, GraSP outperforms ReAct, Reflexion, ExpeL, and flat skill baselines in every configuration, improving reward by up to +19 points over the strongest baseline while cutting environment steps by up to 41%. GraSP's advantage grows with task complexity and is robust to both skill over-retrieval and quality degradation, confirming that structured orchestration -- not larger skill libraries -- is the key to reliable agent execution.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。